Histopathological differences between prostate cancer foci that are detected and missed using multiparametric magnetic resonance imaging in Korean patients

Abstract

Abbreviations & Acronyms BPH = benign prostatic hyperplasia CI = confidence interval Cho = choline Cit = citrate Cr = creatinine DWI = diffusion weighted imaging FSE = fast spin-echo GS = Gleason score HE = hematoxylin-eosin mMRI = multiparametric magnetic resonance imaging MRI = magnetic resonance imaging MRS = magnetic resonance spectroscopy OR = odds ratio PCa = prostate cancer PSA = prostate-specific antigen ROI = region of interest T2WI = T2-weighted imaging Objectives: We investigated whether multiparametric magnetic resonance imaging is appropriate to localize prostate cancer foci in Koreans. Methods: A total of 141 prostate cancer foci in 115 prostate specimens from patients who had undergone radical prostatectomy with preoperative 3 Tesla multiparametric magnetic resonance imaging including T2-weighted imaging, diffusion weighted imaging and magnetic resonance spectroscopy. Differences in the histopathological findings between detected and undetected prostate cancer foci on multiparametric magnetic resonance imaging were investigated. Results: The mean tumor size was 1.9 cm, and 31.9%, 48.9%, and 19.9% of the patients had Gleason scores of 6, 7, or ≥8, respectively. The detection rates of prostate cancer foci were 54.6%, 57.4%, 55.3%, and 45.4% on multiparametric magnetic resonance imaging, T2-weighted imaging, diffusion weighted imaging, and magnetic resonance spectroscopy, respectively. On multivariate analysis, tumor size ≥1.5 cm (odds ratio 3.1; 95% confidence interval 1.31-7.49), Gleason score >7 (4 + 3; odds ratio 2.9; 95% confidence interval 1.05-8.05), and a malignant epithelium/stroma ratio of ≥60% (odds ratio 2.9; 95% confidence interval 1.14-7.20) were significant independent predictors of prostate cancer foci detection on multiparametric magnetic resonance imaging and diffusion weighted imaging. In a multivariate linear model analysis, the apparent diffusion coefficient value was inversely associated with maximum tumor diameter (β = −0.242, P < 0.05), Gleason score (β = −0.234, P < 0.05) and high malignant epithelium/stroma ratio (β = −0.229, P < 0.05). Conclusions: Distinct histological differences between prostate cancer foci that were detected and missed by multiparametric magnetic resonance imaging can be identified. Despite limitations, multiparametric magnetic resonance imaging seems useful for determining prostate cancer in Korean patients, particularly with Gleason score >7 and tumor diameter >1.5 cm.